130 THE ROYAL SOCIETY OF CANADA 
INTRODUCTION. 
If a discrepancy is found between a trustworthy and an untrust- 
worthy record, the conflicting statement of the untrustworthy 
record is disregarded. This practice is based on the correct principle 
of ordinary intercourse that the word of an untrustworthy person has 
no weight, especially against one who is trustworthy. Where the discrep- 
ancy, however, occurs between two trustworthy records, it is ascribed 
to a one-sided stand-point or to unintentional error in one or other of 
the writers. Unless a reasonable harmony can be suggested, the discrep- 
ancy is held to be due to error. 
This is the view of discrepancy usually taken in connection with 
conflicting statements in existing records. The author, on the con- 
trary, has gathered from actual intercourse 26 typical cases in which 
statements that were conflicting were yet true; and these cases are made 
the basis of an experimental study and test of the prevailing view. 
The principal results of the research are those concerning :— 
A. The Origin of Discrepancy. 
B. The Harmonization of Discrepancy. 
C. Discrepancy and Infallibility. 
A. THE ORIGIN OF DISCREPANCY. 
The 26 cases revealed the two following facts which are not recog- 
nized in the prevailing attitude towards discrepancy between trustworthy 
witnesses and records:— 
1. That where two truthful persons make discrepant state- 
ments on account of different stand-points, these stand-points are 
in some cases equally good. 
2. That discrepant statements may be due to a difference of 
stand-point between two inquirers, and in these cases the same 
person, in reporting to the inquirers, must make discrepant state- 
ments in order that each of his statements may be true. 
Illustrations :— 
(1) A farmer, unable to attend an auction himself, bids for a 
property, through an acquaintance, a drover, who secured the place 
at such a low figure that he dishonestly kept it. The farmer will 
